Wavelength division multiplexing (WDM) is a rapidly emerging technology that enables a very significant increase in the aggregate volume of data that can be transmitted over optical fibers. Prior to the use of WDM, most optical fibers were used to unidirectionally carry only a single data channel at one wavelength. The basic concept of WDM is to launch and retrieve multiple data channels in and out, respectively, of an optical fiber. Each data channel is transmitted at a unique wavelength, and the wavelengths are appropriately selected such that the channels do not interfere with each other, and the optical transmission losses of the fiber are low. Today, commercial WDM systems exist that allow for the transmission of 2 to 100 simultaneous data channels.
WDM is a cost-effective method of increasing the volume of data (commonly termed bandwidth) transferred over optical fibers. Alternate competing technologies for increasing bandwidth include the burying of additional fiber optic cable or increasing the optical transmission rate over optical fiber. The burying of additional fiber optic cable is quite costly as it is presently on the order of $15,000 to $40,000 per kilometer. Increasing the optical transmission rate is limited by the speed and economy of the electronics surrounding the fiber optic system. One of the primary strategies for electronically increasing bandwidth has been to use time division multiplexing (TDM), which groups or multiplexes multiple lower rate electronic data channels together into a single very high rate channel. This technology has for the past 20 years been very effective for increasing bandwidth. However, it is now increasingly difficult to improve transmission speeds, both from a technological and an economical standpoint. WDM offers the potential of both an economical and technological solution to increasing bandwidth by using many parallel channels. Further, WDM is complimentary to TDM. That is, WDM can allow many simultaneous high transmission rate TDM channels to be passed over a single optical fiber.
The use of WDM to increase bandwidth requires two basic devices that are conceptually symmetrical. The first device is a wavelength division multiplexer. This device takes multiple beams, each with discrete wavelengths that are initially spatially separated in space, and provides a means for spatially combining all of the different wavelength beams into a single polychromatic beam suitable for launching into an optical fiber. The multiplexer may be a completely passive optical device or may include electronics that control or monitor the performance of the multiplexer. The input to the multiplexer is typically accomplished with optical fibers, although laser diodes or other optical sources may also be employed. As mentioned above, the output from the multiplexer is a single polychromatic beam which is typically directed into an optical fiber.
The second device for WDM is a wavelength division demultiplexer. This device is functionally the opposite of the wavelength division multiplexer. That is, the wavelength division demultiplexer receives a polychromatic beam from an optical fiber and provides a means of spatially separating the different wavelengths of the polychromatic beam. The output from the demultiplexer is a plurality of monochromatic beams which are typically directed into a corresponding plurality of optical fibers or photodetectors.
During the past 20 years, various types of WDMs have been proposed and demonstrated. For example, (1) W. J. Tomlinson, Applied Optics, Vol. 16, No. 8, pp. 2180-2194 (August 1977); (2) A. C. Livanos et al., Applied Physics Letters, Vol. 30, No. 10, pp. 519-521 (May 15, 1977); (3) H. Ishio et al., Journal of Lightwave Technology, Vol. 2, No. 4, pp. 448-463 (August 1984); (4) H. Obara et al., Electronics Letters, Vol. 28, No. 13, pp. 1268-1270 (Jun. 18, 1992); (5) A. E. Willner et al., IEEE Photonics Technology Letters, Vol. 5, No. 7, pp. 838-841 (July 1993); and (6) Y. T. Huang et al., Optical Letters, Vol. 17, No. 22, pp. 1629-1631 (Nov. 15, 1992), all disclose some form of WDM device and/or method. However, most of the WDM devices and/or methods disclosed in the above-listed publications are classical optics-based WDM approaches which employ very basic lenses that are adequate only for use with multimode optical fibers and are inadequate for use with single mode optical fibers because the core diameter of a single mode optical fiber (i.e., typically 8 .mu.m) is much smaller than the core diameter of a multimode optical fiber (i.e., typically 62.5 .mu.m). That is, due to the very basic lenses employed therein, WDM devices incorporating the principles set forth in the classical optics-based WDM approaches disclosed in the above-listed publications are unable to receive and transmit optical beams from and to single mode optical fibers, respectively, without incurring unacceptable amounts of insertion loss and channel crosstalk. These unacceptable levels of insertion loss and channel crosstalk are largely due to the inadequate imaging capabilities of these very basic lenses, which are typically formed of standard optical glass materials.
One proposed solution to the above-described optical imaging problem has been to add additional lenses formed of standard optical glass materials to WDM devices, thereby resulting in WDM devices having doublet, triplet, and even higher number lens configurations. By adding these additional lenses to WDM devices, wherein the added lenses typically have alternating high and low refraction indexes, aberrations caused mainly by the spherical nature of the lenses are effectively canceled out. However, an increased cost is associated with adding these additional lenses due to the direct cost of the additional lenses, as well as the indirect costs associated with the increased complexity and resulting decreased manufacturability of WDM devices having multiple lenses.
Another proposed solution to the above-described optical imaging problem has been to use gradient refractive index lenses (e.g., Gradium lenses) in WDM devices. The use of these gradient refractive index lenses results in a significant improvement in the quality of the imaging system within WDM devices. However, costs associated with manufacturing these gradient refractive index lenses is significantly greater than the costs associated with manufacturing standard homogeneous refractive index lenses, despite the fact that both are typically formed of standard optical glass materials.
In view of the foregoing, there remains a real need for a WDM device which possesses or allows for all the characteristics of: low cost, component integration, environmental and thermal stability, low channel crosstalk, low channel signal loss, ease of interfacing, large number of channels, and narrow channel spacing. Accordingly, it would be desirable to provide a WDM device which overcomes the above-described inadequacies and shortcomings, while possessing or allowing for all of the above-stated characteristics.